Re: [PATCH v2] tests/tcg/s390x: Cleanup of mie3 tests.

On 3/1/22 10:22, David Miller wrote: However the constraint must be wrong there. Sorry about split message. On Tue, Mar 1, 2022 at 3:21 PM David Miller > wrote: I used #define Fi3(S, ASM) uint64_t S(uint64_t a, uint64_t b, uint64_t c) \ {

Re: [PATCH v2] tests/tcg/s390x: Cleanup of mie3 tests.

However the constraint must be wrong there. Sorry about split message. On Tue, Mar 1, 2022 at 3:21 PM David Miller wrote: > I used > > > #define Fi3(S, ASM) uint64_t S(uint64_t a, uint64_t b, uint64_t c) \ > { \ > uint64_t res = 0; \ > asm volatile ( \ >

Re: [PATCH v2] tests/tcg/s390x: Cleanup of mie3 tests.

I used #define Fi3(S, ASM) uint64_t S(uint64_t a, uint64_t b, uint64_t c) \ { \ uint64_t res = 0; \ asm volatile ( \ "ltgr %[c], %[c]\n" \ ASM \ "stg %[c], %[res] " \ : [res] "=" (res) \ : [a] "r" (a), \ [b] "r"

Re: [PATCH v2] tests/tcg/s390x: Cleanup of mie3 tests.

On 3/1/22 09:59, David Miller wrote: +{ \ +uint64_t res = 0; \ +asm volatile ( \ +"lg %%r0, %[c]\n" \ +"ltgr %%r0, %%r0\n" \ +ASM \ +"stg %%r0, %[res] " \ +: [res] "=m" (res) \ +: [a] "r" (a), \ + [b] "r"

[PATCH v2] tests/tcg/s390x: Cleanup of mie3 tests.

Adds clobbers and merges remaining separate asm statements. v1 -> v2: * Corrected side in rebase conflict, removing older code. Signed-off-by: David Miller --- tests/tcg/s390x/mie3-compl.c | 18 +++- tests/tcg/s390x/mie3-mvcrl.c | 12 +++ tests/tcg/s390x/mie3-sel.c | 41